ns-rss-spider 0.0.15 → 0.0.17

package/dist/cjs/parse.d.ts

@@ -7,4 +7,7 @@ export declare function parseRss(name: string, feed: string, server?: ServerInfo
     link: string;
     description: string;
     pubDate: Date;
+    author?: string | undefined;
+    source?: string | undefined;
+    categories?: string | undefined;
 }[]>;

package/dist/cjs/parseContent/generateSsrContent.d.ts

@@ -0,0 +1,8 @@
+/**
+ * 从 content 解析出 app content
+ *
+ * 主要包括：
+ * 1. 图片 placeholder 处理
+ */
+import { RichArticleImage } from "../types";
+export declare function generateSsrContent(content: string, images: RichArticleImage[]): string;

package/dist/cjs/parseContent/generateSsrContent.js

@@ -0,0 +1,48 @@
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+
+// src/parseContent/generateSsrContent.ts
+var generateSsrContent_exports = {};
+__export(generateSsrContent_exports, {
+  generateSsrContent: () => generateSsrContent
+});
+module.exports = __toCommonJS(generateSsrContent_exports);
+var import_cheerio = require("cheerio");
+var import_html_entities = require("html-entities");
+function generateSsrContent(content, images) {
+  const $ = (0, import_cheerio.load)(content, null, false);
+  $("img").each((_, img) => {
+    const src = $(img).attr("src");
+    if (src) {
+      const meta = images.find((i) => i.url === src);
+      const $pl = $(`<span data-img-placeholder></span>`);
+      $pl.attr({
+        "data-w": (meta == null ? void 0 : meta.width) ? `${meta.width}` : "",
+        "data-h": (meta == null ? void 0 : meta.height) ? `${meta.height}` : "",
+        "data-src": src
+      });
+      $pl.attr("style", `display: inline-block;width: ${(meta == null ? void 0 : meta.width) || 1}px; height: ${(meta == null ? void 0 : meta.height) || 1}px`);
+      $(img).replaceWith($pl);
+    }
+  });
+  return (0, import_html_entities.decode)($.html()).trim();
+}
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  generateSsrContent
+});

package/dist/cjs/parseContent/getBasicFromItem.d.ts

@@ -0,0 +1,2 @@
+import { SimpleArticle, RssItem } from "../types";
+export declare function getBasicFromItem(item: RssItem): Pick<SimpleArticle, "guid" | "title" | "link" | "description" | "pubDate" | "author" | "source" | "categories">;

package/dist/cjs/parseContent/getBasicFromItem.js

@@ -0,0 +1,41 @@
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+
+// src/parseContent/getBasicFromItem.ts
+var getBasicFromItem_exports = {};
+__export(getBasicFromItem_exports, {
+  getBasicFromItem: () => getBasicFromItem
+});
+module.exports = __toCommonJS(getBasicFromItem_exports);
+function getBasicFromItem(item) {
+  var _a, _b;
+  return {
+    guid: item.guid,
+    title: item.title,
+    link: item.link,
+    description: item.summary || ((_a = item.contentSnippet) == null ? void 0 : _a.trim()) || "",
+    pubDate: new Date(item.pubDate),
+    author: item.author || item.creator || "",
+    source: item.source || "",
+    categories: ((_b = item.categories) == null ? void 0 : _b.join(",")) || ""
+  };
+}
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  getBasicFromItem
+});

package/dist/cjs/parseContent/index.d.ts

@@ -1,5 +1,4 @@
-import { Item as RssItem } from "rss-parser";
-import { Article, ParseStrategy } from "../types";
+import { Article, ParseStrategy, RssItem } from "../types";
 /**
  * 解析文章内容
  * 1. 获取 html 片段。(rss、http、playwright)

package/dist/cjs/parseContent/index.js

@@ -37,6 +37,7 @@ var cheerio = __toESM(require("cheerio"));
 var import_stripeHtml = require("./stripeHtml");
 var import_parseContent = require("./parseContent");
 var import_zx = require("zx");
+var import_getBasicFromItem = require("./getBasicFromItem");
 async function parseContent(item, strategy) {
   var _a, _b;
   let element;
@@ -55,11 +56,8 @@ async function parseContent(item, strategy) {
   if (!(strategy == null ? void 0 : strategy.parse)) {
     return {
       type: "simple",
-      guid: item.guid,
-      title: item.title,
-      link: item.link,
-      description: ((_b = element.html()) == null ? void 0 : _b.trim()) || "",
-      pubDate: new Date(item.pubDate)
+      ...(0, import_getBasicFromItem.getBasicFromItem)(item),
+      content: ((_b = element.html()) == null ? void 0 : _b.trim()) || ""
     };
   }
   console.log(import_zx.chalk.green("正在解析文章内容"));

package/dist/cjs/parseContent/parseContent.d.ts

@@ -1,4 +1,3 @@
-import { Item as RssItem } from "rss-parser";
-import type { ParseStrategy, RichArticle } from "../types";
+import type { ParseStrategy, RichArticle, RssItem } from "../types";
 import { AnyNode, Cheerio, CheerioAPI } from "cheerio";
 export declare function parseContent($: CheerioAPI, $element: Cheerio<AnyNode>, item: RssItem, strategy?: ParseStrategy): Promise<RichArticle>;

package/dist/cjs/parseContent/parseContent.js

@@ -36,8 +36,9 @@ var import_utils = require("./utils");
 var import_probe_image_size = __toESM(require("probe-image-size"));
 var import_zx = require("zx");
 var import_html_entities = require("html-entities");
+var import_generateSsrContent = require("./generateSsrContent");
+var import_getBasicFromItem = require("./getBasicFromItem");
 async function parseContent($, $element, item, strategy) {
-  var _a;
   const srcs = [];
   const images = [];
   (0, import_utils.walk_the_DOM)($, $element, (node) => {
@@ -82,7 +83,6 @@ async function parseContent($, $element, item, strategy) {
   $($element).find("img").each((_, img) => {
     const src = $(img).attr("src");
     if (src) {
-      $(img).removeAttr("src");
       $(img).attr("data-src", src);
       const meta = images.find((i) => i.url === src);
       if (meta == null ? void 0 : meta.width) {
@@ -93,14 +93,12 @@ async function parseContent($, $element, item, strategy) {
       }
     }
   });
+  const content = (0, import_html_entities.decode)($element.html().trim());
   return {
     type: "rich",
-    guid: item.guid,
-    title: item.title,
-    link: item.link,
-    description: ((_a = item.contentSnippet) == null ? void 0 : _a.trim()) || "",
+    ...(0, import_getBasicFromItem.getBasicFromItem)(item),
     content: (0, import_html_entities.decode)($element.html().trim()),
-    pubDate: new Date(item.pubDate),
+    ssrContent: (0, import_generateSsrContent.generateSsrContent)(content, images),
     images
   };
 }

package/dist/cjs/types.d.ts

@@ -1,5 +1,9 @@
 import type { AnyNode, Cheerio, CheerioAPI } from 'cheerio';
 import { Item } from 'rss-parser';
+export type RssItem = Item & {
+    source?: string;
+    author?: string;
+};
 export interface ParseStrategy {
     parse: boolean;
     fetcher?: 'http' | 'playwright';
@@ -12,7 +16,11 @@ export interface SimpleArticle {
     title: string;
     link: string;
     description: string;
+    content: string;
     pubDate: Date;
+    author?: string;
+    source?: string;
+    categories?: string;
 }
 export interface RichArticleImage {
     url: string;
@@ -26,7 +34,7 @@ export interface RichArticleImage {
 }
 export interface RichArticle extends Omit<SimpleArticle, 'type'> {
     type: 'rich';
-    content: string;
+    ssrContent: string;
     images: RichArticleImage[];
 }
 export type Article = SimpleArticle | RichArticle;

package/dist/cjs/upload/index.d.ts

@@ -18,4 +18,7 @@ export declare function uploadContent({ app, getUploadApi, items, }: {
     link: string;
     description: string;
     pubDate: Date;
+    author?: string | undefined;
+    source?: string | undefined;
+    categories?: string | undefined;
 }[]>;

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "ns-rss-spider",
-  "version": "0.0.15",
+  "version": "0.0.17",
   "description": "",
   "main": "dist/cjs/index.js",
   "types": "dist/cjs/index.d.ts",